STIX Simulated Data Generation

• Before running anything on the repo:

pip install -r requirements.txt

• Generate a dataset generate_ds.py of 1 to 4 number of sources (in equal amount) and save it as .npz:

$ python generate_ds.py -p=out_data -s=40000 -t=test -n=13

13 was the chosen seed for the test set, other choices for this project include validation data -> 4, training data -> 33.

Citation

If this repository proves useful for your research, please cite below work in which the synthetic data generated via this repository was introduced.

@article{SelcukSimsek2025,
 author    = {Selcuk-Simsek, Merve and Massa, Paolo and Xiao, Hualin and Krucker, S{\"a}m and Csillaghy, Andr{\'e}},
 title     = {Fourier convolutional decoder: reconstructing solar flare images via deep learning},
 journal   = {Neural Computing and Applications},
 pages     = {1--32},
 year      = {2025},
 publisher = {Springer}
 doi       = {10.1007/s00521-025-11283-6},
 url       = {https://doi.org/10.1007/s00521-025-11283-6}
}

@article{massa_stix_2023,
   title = {The {STIX} {Imaging} {Concept}},
   volume = {298},
   issn = {1573-093X},
   url = {https://doi.org/10.1007/s11207-023-02205-7},
   doi = {10.1007/s11207-023-02205-7},
   number = {10},
   journal = {Solar Physics},
   author = {Massa, Paolo and Hurford, Gordon J. and Volpara, Anna and Kuhar, Matej and Battaglia, Andrea F. and Xiao, Hualin and Casadei, Diego and Perracchione, Emma and Garbarino, Sara and Guastavino, Sabrina and Collier, Hannah and Dickson, Ewan C. M. and Emslie, A. Gordon and Ryan, Daniel F. and Maloney, Shane A. and Schuller, Frederic and Warmuth, Alexander and Massone, Anna Maria and Benvenuto, Federico and Piana, Michele and Krucker, Säm},
   month = oct,
   year = {2023},
   keywords = {Flares, Instrumentation and data management, Spectrum, X-ray},
   pages = {114},
}

Note

This repository uses a previous version of the original STIX data simulator for which you can find the original and up-to-date version here.